The Time Division Multiplexing (TDM) based Passive Optical Network (PON) system had been world widely deployed because it is the most cost-effective way to implement fiber to home service. However, rouge Optical Network Unit (ONU) may lead part (or whole) of PON network service broken.
For the design of a TDM PON, the Optical Line Terminal (OLT) talks to OUNs in a broadcast manner over the downstream wavelength on optical fiber. And each ONU talks to the OLT within a given time slot allocated by the OLT over upstream wavelength (normally, upstream and downstream wavelength is different on the same trunk fiber). If a rogue ONU intentionally or unintentionally transmits optical signals to the OLT in the time slot which is allocated to other ONUs, the upstream service transmitted by other ONUs over this trunk fiber will be disturbed seriously.
“Rogue ONU” typically refers to an ONU which is transmitting in a manner not consistent with parameters specified in the standard. In the actual PON, it is an important issue to be solved, as the rogue ONU prevents the normal communications between the ONUs and the OLT. The rogue behavior of the rogue ONU can be caused by many reasons such as MAC error, software error, transmitter error, and so on. The most common case happens both in the TDM-PON and in the TWDM-PON is that the rogue ONU transmits its upstream signal in wrong time slots, as shown in FIG. 1. Normally, the failed transmitter turns on ahead of schedule because of the software or hardware error; or it turns off later than the schedule because of the component aging or other reasons. Therefore, the upstream bursts from other ONUs which are adjacent to the rogue ONU will be disturbed and cannot be received correctly by the OLT. However, determining a rogue ONU can be difficult since the affected ONUs are not always the ONUs which cause the disturbance.
In ITU and CCSA, typically there are several solutions to determine a rogue ONU.
Solution 1: powering off/on each ONU one by one till a rogue ONU is determined. However, this solution has the following disadvantages: 1) It is a time-consuming process since the OLT needs to check each ONUs' status one by one; 2) During the process of rogue ONU determination, the normal communication between the ONUs and the OLT may be interrupted, a long-time service interruption is unacceptable to the users; 3) Most Rogue ONUs are not always in rogue status and they may come back in a normal status after power/laser reset or fiber pull/plug; 4) This solution cannot guarantee the high reliability of a PON system.
Solution 2: calculating the error rate and link broken times for each ONU and recording the Dynamic Bandwidth Allocation (DBA) history in the OLT. Based on the DBA history, ONUs (with high error rate), which are impacted by adjacent rogue ONU, could be found out. In theory, all impacted normal ONUs should be disturbed by the same rogue ONU (with low error rate). By this way, the rogue ONU which is adjacent to and impact those normal ONU's (with high error rate) can be located. But this solution is not so pragmatic, because most OLTs of a PON do not record its DBA history and the adjacent information of the time slot is not available.
Therefore, design a solution in a practical system for identifying the rogue ONU quickly and remotely without service interruption is very beneficial.